@yrathi as far as I can tell there are currently no tests of NODDI, but we would like to add some to avoid regressions. If you have any "known correct" result data, that could be very helpful.
Otherwise the simplest thing would be to add baseline generated from some early commit after noddi was added with a variety of permutations on existing test data (and of course make sure the output is currently the same).
cc @ljod @sbouix
@yrathi as far as I can tell there are currently no tests of NODDI, but we would like to add some to avoid regressions. If you have any "known correct" result data, that could be very helpful.
Otherwise the simplest thing would be to add baseline generated from some early commit after noddi was added with a variety of permutations on existing test data (and of course make sure the output is currently the same).
cc @ljod @sbouix